Abstract:We study online prediction for marginally stable, partially observed linear dynamical systems under nonstochastic disturbances. Our objective is to minimize the cumulative squared prediction loss and compete with the best-in-hindsight Luenberger predictor. Standard online learning methods typically rely on bounded domains/gradients, and thus their guarantees may fail to deal with potentially unbounded trajectories in marginally stable systems. In this paper, we introduce an unconstrained online least squares method that stabilizes the learning process via tailored predictive hints. With model knowledge, we prove that hints constructed from any stabilizing Luenberger predictor render the hint residuals uniformly bounded, achieving logarithmic regret despite unbounded trajectory growth. We also discuss model-free prediction and introduce a simple universal hint for symmetric systems, under which logarithmic regret is maintained without model knowledge. Our results provide an adaptive, instance-wise optimal online predictor compared to classical fixed-gain observers under nonstochastic disturbances.
| Subjects: | Machine Learning (cs.LG); Systems and Control (eess.SY); Optimization and Control (math.OC) |
| Cite as: | arXiv:2605.04364 [cs.LG] |
| (or arXiv:2605.04364v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.04364 arXiv-issued DOI via DataCite (pending registration) |
From: Chih-Fan Pai [view email]
[v1]
Wed, 6 May 2026 00:01:54 UTC (728 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。